Visualization Interface of Comment Streams from Social Network Services

This paper focus on generating a clear visualization interface of comment streams from Social Network Services. For each message on social network services, clients can express their conclusions by sending, giving a like, and leaving comments on it. The amount of comment is expansive, as well as the generation rate is also high. Clients may want to get a concise comprehension of a comment stream without analyzing the entire comment list, so we attempt to cluster comments with comparable substance together to produce a brief summary of message. A bunch form of short content rundown calculation (Batch STS) is first presented. Since particular clients will ask for the summary at any minute, a novel incremental grouping calculation called IncreSTS calculation can incrementally refresh comments with most recent comments continuously. IncreSTS has the benefits of high productivity, high versatility, and better dealing with anomalies. This paper uses two algorithms Batch STS and IncreSTS for efficient clustering of comments and focus on language processing by implementing a 4 Gram generation of comments in order to take semantic details of comments into account. Various NLP techniques are implemented such as stemming, redundant character removal and n gram generation for efficient short text summarization. Finally a clear visualization of the summary is generated.